Synthesis of epi-oxetin via a serine-derived 2-methyleneoxetane

Article abstract of DOI:10.1021/jo7018762

(Chemical Equation Presented) The unique reactivity of 2-methyleneoxetanes and 1,5-dioxaspiro[3.2]hexanes has been exploited for the synthesis of epi-oxetin (26), an oxetane-containing β-amino acid. While the preparation of the natural product oxetin (1) was the original goal, the unexpected diastereoselectivity of an precedented reduction provided the epi-oxetin framework. The methodology described herein should be amenable for the preparation of oxetin with a change in nitrogen protection.

Full text of DOI:10.1021/jo7018762

Synthesis of epi-Oxetin via a Serine-Derived 2-Methyleneoxetane
Marisa L. Blauvelt and Amy R. Howell*
Department of Chemistry, UniVersity of Connecticut, Storrs, Connecticut 06269-3060
amy.howell@uconn.edu
ReceiVed August 27, 2007
The unique reactivity of 2-methyleneoxetanes and 1,5-dioxaspiro[3.2]hexanes has been exploited for the
synthesis of epi-oxetin (26), an oxetane-containing â-amino acid. While the preparation of the natural
product oxetin (1) was the original goal, the unexpected diastereoselectivity of an precedented reduction
provided the epi-oxetin framework. The methodology described herein should be amenable for the
preparation of oxetin with a change in nitrogen protection.
Introduction
In 1984 â-amino acid 1 (Figure 1) was isolated from the
culture filtrate of Streptomyces sp. OM-2317, a bacteria orig-
inating in a Japanese soil sample.1 Due to its oxetane core, this
novel compound was named oxetin. Structural determination
via X-ray crystallography assigned an absolute stereochemistry
of 2R,3S, which was later confirmed by the synthesis of the
natural product by Omura and co-workers.2 Their synthesis of
oxetin, starting from D-glucose, provided oxetin and its three
stereoisomers due to the nonstereoselective nature of several
key steps.
FIGURE 1. Structure of oxetin.
Biological testing of oxetin revealed both antibiotic and
herbicidal activity. Oxetin inhibited glutamine synthetase from
Bacillus subtilis and spinach (Spinacia oleracea). Other non-
competitive glutamine antimetabolites, such as tabtoxin, are
known to harm plants via inhibition of glutamine synthetase.3
Moreover, oxetin exhibited antibacterial activity against Bacillus
subtilis, Piricularia oryzae, and other microorganisms. Oxetin’s
three stereoisomers were found to be inactive.
FIGURE 2. Proposed strategy for the synthesis of oxetin via
2-methyleneoxetane 4.
Results and Discussion
Our initial strategy for the synthesis of oxetin is shown in
Figure 2. The preparation of a variety of serine lactones (e.g.,
P ) Boc,7 Boc and allyl,8 pthalimide,9 CBz10) has been reported.
We have previously reported the synthesis of serine-derived 1,5-
dioxaspiro[3.2]hexane 5 (P ) Boc) from the corresponding
A more concise synthesis utilizing the Paterno-Bu¨chi reac-
tion by Bach and Schro¨der was published in 1997.4 Although
this synthesis was shorter, it was racemic. Due to the oxetane
core structure, we were interested in synthesizing oxetin via a
2-methyleneoxetane, which we have shown to be readily
accessible from â-lactones.5,6
(5) Dollinger, L. M.; Howell, A. R. J. Org. Chem. 1996, 61, 7248.
(6) Dollinger, L. M.; Ndakala, A. J.; Hashemzadeh, M.; Wang, G.; Wang,
Y.; Martinez, I.; Arcari, J. T.; Galluzzo, D. J.; Howell, A. R.; Rheingold,
A. L.; Figuero, J. S. J. Org. Chem. 1999, 64, 7074.
(7) Pansare, S. V.; Arnold, L. D.; Vederas, J. C. Org. Synth. 1992, 70,
10.
* Address correspondence to this author. Phone: 860-486-3460. Fax: 860-
486-2981.
(1) Omura, S.; Murata, M.; Imamura, N.; Iwai, Y.; Tanaka, H.; Furusaki,
A.; Matsumoto, T. J. Antibiot. 1984, 37, 1324.
(8) Soucy, F.; Wernic, D.; Beaulieu, P. J. Chem. Soc., Perkin Trans. 1
1991, 2885.
(2) Kawahata, Y.; Takatsuto, S.; Ikekawa, N.; Murata, M.; Omura, S.
Chem. Pharm. Bull. 1986, 34, 3102.
(3) Sinden, S. L.; Durbin, R. D. Nature 1968, 219, 379.
(4) Bach, T.; Schro¨der, J. Liebigs Ann./Recl. 1997, 2265.
(9) Lall, M. S.; Ramtohul, Y. K.; James, M. N. G.; Vederas, J. C. J.
Org. Chem. 2002, 67, 1536.
(10) Marinez, E. R.; Salmassian, E. K.; Lau, T. T.; Gutierrez, C. G. J.
Org. Chem. 1996, 61, 3548.
10.1021/jo7018762 CCC: $40.75 © 2008 American Chemical Society
Published on Web 12/20/2007
J. Org. Chem. 2008, 73, 517-521
517

Blauvelt and Howell
SCHEME 1
FIGURE 3. Formation of an aluminum-stabilized oxonium ion.
TABLE 1. Attempted Reduction Conditions for 17
entry
reducing agent
equiv
ratio of 18:19a
2-methyleneoxetane 4 (P ) Boc).11 On the basis of our previous
studies on the reactivity of dioxaspirohexanes with DIBAL-
H,12 we anticipated that oxetane intact product 6 could be
secured. Oxidation to the corresponding carboxylic acid and a
final deprotection would yield oxetin (1).
Lactone 7 was prepared under the Mitsunobu conditions given
in Scheme 1.7 Methylenation with the Petasis reagent to form
8, however, proceeded in modest yield despite numerous
modifications in reaction conditions. In contrast, epoxidation
with anhydrous, acetone-free dimethyldioxirane produced 1,5-
dioxaspiro[3.2]hexane 9 in quantitative yield.
1
2
3
4
5
6
DIBAL-H in hexanes
DIBAL-H in hexanes
DIBAL-H in hexanes
DIBAL-H in hexanes
DIBAL-H in DCM
TMSCl, TEA,
1.0
2.5
5
2:1
2:1
1.8:1
1.7:1
10
2.5
1.2b
2.6:1
DIBAL-H
7
8
9
Mg(OTf)2, DIBAL-H
Mg(OTf)2, Et3SiH
Zn(BH4)2
1.0
1.0
1.1
1.0
3:1
10
NaBH3CN
a NMR ratios of crude reaction mixtures. b A similar outcome was seen
with either 2 or 3 equiv of TMSCl.
In 1999, we reported on the dichotomy of reactivity of 1,5-
dioxaspiro[3.2]hexanes.12 Although many nucleophiles gave
R-functionalized-â-hydroxyketones, Lewis acidic reagents pro-
vided oxetane-intact products. Specifically, the reaction of 11
with DIBAL-H afforded predominantly oxetane 13 (Figure 3).
Formation of 13 was rationalized by the coordination of the
Lewis acid to the oxirane oxygen and participation of oxonium
ion 12. Subsequent hydride delivery was presumably favored
on the less hindered face, leading to an 8:1 mixture of
diastereomers.
SCHEME 2
We anticipated that the reduction of N-Boc-1,5-dioxaspiro-
[3.2]hexane 9 would proceed with similar facial selectivity,
providing the desired 2R,3S stereochemistry. After numerous
attempts at the DIBAL-H reduction, no trace of alcohol 10 was
observed. We speculated that the Boc carbonyl could be
participating in stabilization of reaction intermediates. The
failure of the reduction, coupled with the low-yielding meth-
ylenation step, prompted us to select an alternative nitrogen
protecting group.
Campagne and Ghosez reported on an improved efficiency
of RCM reactions of Boc-benzyl protected vinyl glycine
derivatives over their Boc-protected counterparts.13 Furthermore,
these RCM reactions proceeded with even higher yields with
the utilization of triphenylmethyl protection. Due to the potential
difficulties and additional steps required to add and remove two
nitrogen protecting groups, we examined N-trityl serine for the
synthesis of oxetin.
N-Trityl serine (14), obtained in one step from L-serine,
lactonized efficiently in the presence of the coupling agent BOP
to yield 15.14 Methylenation of the N-trityl â-lactone 15 to
methyleneoxetane 16 was both cleaner and higher yielding than
was observed with the corresponding Boc-protected lactone 7,
and oxidation provided 17 (as a 2:1 mixture of diastereomers).
When dioxaspirohexane 17 was treated with 1 equiv of
DIBAL-H (Table 1, entry 1), two diastereomers (2:1 ratio)
resulted, but the isolated yield was low (∼30%). With 2.5 equiv
of DIBAL-H reduction of 17 gave diastereomers 18 and 19
(Table 1, entry 2), also as a 2:1 mixture, in a combined 65%
yield. NOESY experiments provided the relative configuration
of the diastereomers. Unexpectedly, minor isomer 19 clearly
exhibited NOE between the hydrogens H1 and H2 (see Scheme
2 and the Supporting Information), while major isomer 18 did
not. Although the overall reduction yield was good, the
diastereoselectivity was low and favored the undesired isomer.
Subsequently, a variety of reduction conditions were examined,
as shown in Table 1, and none provided 19 as the major product
(Table 1).
Reasoning that the DIBAL-H might be complexing with the
nitrogen lone pair and subsequently delivering the hydride to
the same face, a large excess of DIBAL-H (entries 3 and 4)
was added to see if the excess reducing agent might compete
to provide hydride delivery to the less hindered side. Only a
slight improvement in the desired stereoselectivity was seen.
In a related strategy, a Lewis acid was first added to interact
with the Lewis basic nitrogen prior to the addition of DIBAL-H
with the goal of decreasing internal hydride transfer. With prior
addition of 1.2 to 3 equiv of TMSCl in the presence of
triethylamine (TEA), diastereoselectivity improved in the wrong
direction! The addition of Mg(OTf)2, however, did not give
oxetane intact products.
(11) Ndakala, A. J.; Howell, A. R. J. Org. Chem. 1998, 63, 6098.
(12) Howell, A. R.; Ndakala, A. J. Org. Lett. 1999, 1, 825.
(13) Campagne, J.; Ghosez, L. Tetrahedron Lett. 1998, 39, 6175.
(14) Sliedregt, K. M.; Schouten, A.; Kroon, J.; Liskamp, R. M. J.
Tetrahedron Lett. 1996, 37, 4237.
Other reducing agents were also examined. Zinc borohydride
gave 18 and 19 with a greater ratio of the undesired diastere-
omer. Dioxaspirohexane 17 was consumed in the presence of
sodium cyanoborohydride and triethylsilane, but no alcohol
518 J. Org. Chem., Vol. 73, No. 2, 2008

Synthesis of epi-Oxetin
resulted. Surprisingly, a similar result was seen when DIBAL-H
in dichloromethane (rather than hexanes) was used.
Conclusions
Herein we have described the synthesis of epi-oxetin (26) in
10 steps, starting from the natural amino acid L-serine, via
2-methyleneoxetane 16. It appears that, due to either the Lewis
basic properties of the nitrogen or inclusionary attributes of the
trityl group, the diastereoselectivity of the key reduction of
dioxaspirohexane 17 proceeded in favor of the undesired alcohol
18. Future work includes employing methodology described
herein with alternative nitrogen protecting groups, including
pthalimide or nitrogen double protection.
It has been reported that triphenylmethanol, as well as other
molecules containing triaryl moieties, can act as inclusionary
compounds.15,16 Triphenylmethanol has extraordinary inclusion
selectivity, allowing such molecules as methanol, DMSO, and
DMF as guests in its crystal lattice. N-Triphenylmethyl moieties
are also reported to act as “spacers”, preventing host molecules
from tightly packing, facilitating the incorporation of guest
molecules into the crystal.17 When extensive efforts to com-
pletely remove ethyl acetate from 18 or 19 failed, it seemed
likely that 18 or 19 could be hosting ethyl acetate as a guest
molecule.
Experimental Section
A sample of 18 was washed with MeOH followed by solvent
removal under vacuum. 1H NMR (in CD3OD) showed that the
ethyl acetate was now replaced with MeOH. From this observa-
tion, we postulated that DIBAL-H could act as a guest in the
aromatic pocket, subsequently delivering the hydride from the
more hindered face. At this point, we decided to pursue the
synthesis of epi-oxetin from 18 in order to develop methodology
that could be later utilized in the formation of oxetin and oxetin
derivatives.
Procedure for the Preparation of Dimethyldioxirane. Dim-
ethyldioxirane (DMDO) was prepared and concentrated as described
by Messegeur.22 NaHCO3 (240 g), acetone (260 mL), and water
(350 mL) were charged into a 3 L round-bottomed flask with a
large magnetic stir bar. The mixture was cooled with an ice bath
and vigorously stirred. Oxone (450 g) was slowly added to the
mixture, and vacuum was applied (80 mmHg). After 10 min, the
ice bath was removed, and the reaction was allowed to warm to rt,
where DMDO in acetone was trapped over 1 h into two receiving
flasks in series maintained at -78 °C. The combined solutions were
diluted with H2O (220 mL), and extracted with cold CH2Cl2 (3 ×
11 mL). The combined organic extracts were neutralized with
phosphate buffer (pH 7, three times equal to the volume of organic
extract), dried (MgSO4), and filtered. The yellow CH2Cl2/DMDO
solution was stored over 4 Å molecular sieves in the freezer. The
concentration of DMDO was determined by 1H NMR and was based
upon the reaction between excess citronellic acid and DMDO.
Concentrations vary from 0.3 to 0.5 M and have been shown to be
>95% accurate.
Direct oxidation of 18 to the carboxylic acid was problematic.
A number of procedures, including PDC/DMF, ruthenium
trichloride hydrate, and sodium periodate, failed to provide 21.
Attempted oxidations to aldehyde 20, using Swern conditions,
PCC, PDC/DCM, and TPAP/NMO, were also fruitless. By 1H
NMR, it was clear that the N-trityl group did not survive any
of the oxidation conditions; therefore, we switched to another
nitrogen protecting group.
Preparation of Dimethyltitanocene.6 MeLi (1.4 M in Et2O, 55
mL) was added dropwise over 30 min to a slurry of titanocene
dichloride (10.0 g, 40.1 mmol) in toluene (80 mL) at -5 °C under
N2. The resulting mixture was then stirred at -5 °C for an additional
hour and then brought to 0 °C. The reaction was then slowly
quenched by the additional of a 6% NH4Cl solution (75 mL). The
aqueous layer was then extracted three times with Et2O (3 × 100
mL). The organic layers were combined and washed with H2O (3
× 50 mL) and brine (3 × 50 mL), then dried (MgSO4), filtered,
and concentrated to approximately one-third the volume of toluene.
1H NMR assay indicated 7.18 g (86%) of dimethyltitanocene. The
deep red solution was diluted to 0.5 M with toluene and stored in
the freezer.
A variety of approaches to trityl cleavage were pursued. While
hydrogenolysis of 18 with Pd black18 resulted in the recovery
of starting material, TFA, Yb(OTf)3,19 and HCl,20 yielded no
oxetane intact products. Protection of 18 by acetic anhydride,
followed by a one-pot TFA trityl deprotection-Boc reprotection,
provided 23 in good yield. Following cleavage of the acetate
group, oxidation to carboxylic acid 25 proceeded modestly using
ruthenium trichloride hydrate and sodium periodate.21 Boc
deprotection using TFA rendered epi-oxetin (26) (Scheme 3).
1H and 13C NMR, as well as the optical rotation of 26, agreed
with literature values.2
N-Tritylserine (14). L-Serine (3.15 g, 30.0 mmol) was suspended
in CH2Cl2 (53 mL) under N2. At rt, TMSCl (13.3 mL, 105 mmol)
was added via syringe, and the mixture was heated at reflux for 1
h. The solution was allowed to cool to rt, and triethylamine (14.7
mL, 105 mmol) in CH2Cl2 (30 mL) was added slowly, whereupon
(15) Weber, E.; Konstantinos, S.; Wierig, A. J.; Goldberg, I. Inclusion
Phenom. Mol. Recognit. 1997, 28, 163.
(16) Weber, E.; Skobridis, K.; Goldberg, I. J. Chem. Soc., Chem
Commun. 1989, 1195.
SCHEME 3
(17) Goldberg, I.; Lin, L. W.; Hart, H. Acta. Crystallogr. 1985, C41,
1535.
(18) Dugave, C.; Menez, A. J. Org. Chem. 1996, 61, 6067 and references
cited therein.
(19) Lu, R. J.; Liu, D.; Giese, R. W. Tetrahedron Lett. 2000, 41, 2817.
(20) Applegate, H. E.; Cimarusti, C. M.; Dolfini, J. E.; Funke, P. T.;
Koster, W. H.; Puar, M. S.; Slusarchyk, W. A.; Young, M. G. J. Org. Chem.
1979, 44, 811.
(21) Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K. B. J.
Org. Chem. 1981, 46, 3936.
(22) Ferrer, M.; Gilbert, M.; Sanchez-Baeza, F.; Messeguer, A. Tetra-
hedron Lett. 1996, 37, 3585.
J. Org. Chem, Vol. 73, No. 2, 2008 519

Blauvelt and Howell
the mixture became difficult to stir. The mixture was then heated
at reflux for 45 min. Upon cooling to 0 °C, MeOH (1.8 mL) in
CH2Cl2 (7.5 mL) was added, followed by the addition of triethy-
lamine (4.2 mL, 30 mmol) and triphenylmethyl chloride (8.37 g,
30.0 mmol) in 2 portions over 15 min. The reaction was allowed
to stir overnight. MeOH (6.0 mL) and triethylamine (4.2 mL) were
then added, and the mixture was stirred for an additional 15 min.
All solvents were removed under vacuum, and the crude acid was
dissolved in EtOAc (150 mL) and washed with a 5% citric acid
solution (3 × 75 mL) and then with H2O (3 × 75 mL). The organic
layer was dried (MgSO4), filtered, and concentrated, yielding a
yellow foam (85%). The crude acid was purified by precipitation
from CHCl3, and the product was isolated as a white solid (6.9 g,
66%):14 1H NMR (400 MHz, CDCl3/CD3OD) δ 7.44 (d, J ) 7.6
Hz, 6H), 7.26 (m, 9H), 3.71(d, J ) 11.1 Hz, 1H), 3.49 (s, 1H),
3.11 (m, 3H), 2.90 (dd, J ) 4.3, 11.1 Hz, 1H); 13C NMR (100
MHz, CDCl3/CD3OD) δ 175.0, 144.7, 128.8, 128.4, 127.3, 72.1,
63.0, 58.7.
(S)-3-(Tritylamino)oxetan-2-one (15). N-Tritylserine (14) (1.75
g, 5.04 mmol) was suspended in CH2Cl2 (25 mL) under N2 at rt.
Triethylamine (2.0 mL, 14 mmol) was added, and the mixture
became homogeneous. Benzotriazol-1-yloxytris(dimethylamino)-
phosphonium hexafluorophosphate (3.12 g, 7.06 mmol) was added
in two portions over 15 min and the solution was stirred for an
additional hour. The reaction was then diluted with H2O (25 mL)
and stirred for an additional 20 min. The layers were separated,
and the aqueous layer was extracted with CH2Cl2 (3 × 25 mL).
The combined organic layers were then dried (MgSO4), filtered,
and concentrated, yielding a yellow solid. The crude solid was
purified using flash column chromatography on silica gel (petroleum
ether/EtOAc 85:15) to give lactone 15 (1.2 g, 73%) as a white solid:
4.25 (m, 2H), 3.12 (m, 1H), 2.81 (d, J ) 8.9 Hz, 1H), 2.70 (d, J )
3.2 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 146.2, 128.5, 128.1,
126.7, 95.1, 72.8, 70.8, 54.0, 50.2; minor diastereomer δ 7.43 (d,
J ) 7.4 Hz, 6H), 7.22 (m, 9H), 4.45-4.25 (m, 2H), 3.58 (t, J )
6.6 Hz, 1H), 2.86 (d, J ) 4.0 Hz, 1H), 2.42 (d, J ) 12.5 Hz, 1H);
13C NMR (100 MHz, CDCl3) δ 145.7, 128.5, 128.1, 126.8, 94.9,
70.1, 56.2, 49.0; MS (EI) m/z 274, 243, 197 (100), 165, 105, 77;
HRMS (FAB) calcd for C23H22NO2 (M+ + H) 344.1651, found
328.1714.
2-Hydroxymethyl-3-tritylaminooxetanes 18 and 19. 3-(Trity-
lamino)-1,5-dioxaspiro[3.2]hexane (17) (0.39 g, 1.2 mmol) was
diluted with dry toluene (25 mL), and the solution was cooled to
-78 °C under N2. A solution of DIBAL-H (1.0 M in hexanes, 2.9
mL, 2.9 mmol) was added dropwise via syringe over 20 min. The
solution was stirred at -78 °C for 1 h. The reaction was quenched
with MeOH (0.5 mL) at -78 °C and slowly brought to 0 °C, where
it was quickly flashed on silica gel (petroleum ether/EtOAc 70:30)
to avoid the formation of aluminum gels. After the evaporation of
all solvents, the resulting oil was repurified on silica gel (petroleum
ether/EtOAc 70:30). Two diastereomers, 18 and 19, were isolated.
The major diastereomer (2S,3S)-2-hydroxymethyl-3-tritylaminoox-
etane (18) was isolated as a white foam (0.20 g, 48%): [R]25D +24.0
1
(c 0.3, CH2Cl2); IR (KBr) 3396 (br), 3056, 2923, 1595 cm-1; H
NMR (400 MHz, CD3OD) δ 7.39 (m, 6H), 7.13 (m, 9H), 4.50 (m,
1H), 3.73 (m, 2H), 3.62 (m, 1H), 3.36 (dd, J ) 2.5, 12.7 Hz, 1H),
3.22 (m, 2H); 13C NMR (100 MHz, CD3OD) 148.0, 129.9, 129.2,
127.6, 93.1, 79.0, 71.6, 64.1, 51.8; MS (EI) m/z 244 (+HC(Ph)3)
(100), 165, 152, 115, 77, 51; HRMS (FAB) calcd for C23H24NO2
(M+ + H) m/z 346.1807, found 346.1792. Minor diastereomer
(2R,3S)-2-hydroxymethyl-3-tritylaminooxetane (19) was isolated as
a cloudy oil (0.056 g, 14%): [R]25 -50.9 (c 1.7, CH2Cl2); IR
D
14 [R]25D -70.4 (c 0.6, CH2Cl2); IR (KBr) 3322, 3063, 1816 cm-1
;
(KBr) 3396 (br), 3056, 2922, 1595 cm-1; H NMR (400 MHz,
1
1H NMR (400 MHz, CDCl3) δ 7.56 (d, J ) 7.4 Hz, 6H), 7.30 (m,
9H), 4.65 (m, 1H), 3.55 (dd, J ) 5.7, 5.7 Hz, 1H), 3.19 (dd, J )
5.1, 5.1 Hz, 1H), 2.76 (d, J ) 11.2 Hz, 1H); 13C NMR (100 MHz,
CDCl3) δ 172.2, 145.3, 128.5, 127.1, 70.9, 70.7, 64.7; MS (EI)
m/z 243 (+CPh3) (100), 228, 165, 77.
CDCl3) δ 7.48 (d, J ) 7.6 Hz, 6H), 7.27 (m, 9H), 4.75 (m, 1H),
4.20 (dd, J ) 7.5, 7.5 Hz, 1H), 4.11 (dd, J ) 6.4, 6.4 Hz, 1H),
3.83 (m, 2H), 3.68 (d, J ) 11.8 Hz, 1H); 13C NMR (100 MHz,
CDCl3) δ 146.5, 128.3, 128.0, 126.7, 87.6, 82.0, 70.4, 63.0, 50.9;
MS (EI) m/z 244 (+HC(Ph)3) (100), 165, 152, 115; HRMS (FAB)
calcd for C23H24NO2 (M+ + H) m/z 346.1807, found 346.1795.
(2S,3S)-Acetic Acid 3-(Tritylamino)oxetan-2-ylmethyl Ester
(22). (2S,3S)-2-Hydroxymethyl-3-tritylaminooxetane (18) (0.21 g,
0.61 mmol) was dissolved in pyridine (5.0 mL) and stirred at rt
under N2. Acetic anhydride (0.10 mL, 0.90 mmol) and a catalytic
amount of DMAP were added. Upon consumption of starting
material (20-40 min), pyridine was removed in vacuo, and the
resulting pale yellow oil was diluted with CH2Cl2 (10 mL). The
solution was washed with 10% aqueous CuSO4 (3 × 25 mL), H2O
(3 × 25 mL), and brine (3 × 25 mL), then dried (MgSO4).
Evaporation of all solvents yielded acetate 22 (160 mg, 71%) as a
(S)-2-Methylene-3-(tritylamino)oxetane (16). (S)-3-(Tritylami-
no)oxetan-2-one (15) (1.50 g, 4.58 mmol) was dissolved in a
solution of dimethyltitanocene (0.5 M in toluene, 23 mL) and further
diluted with toluene (20 mL). The reaction was then heated to
80 °C under N2 in the dark and monitored by TLC for the
disappearance of starting material (2-5 h). The cooled reaction
mixture was added to petroleum ether (3 volumes), and the resulting
suspension was stirred overnight. The orange precipitate was
separated from the red filtrate using a pad of celite. The filtrate
was concentrated in vacuo to approximately one-fourth the volume
of toluene and purified by flash column chromatography on silica
gel (petroleum ether/triethylamine/EtOAc 99:0.5:0.5). Methyl-
clear oil that was used without further purification: [R]25 +19.3
D
eneoxetane 16 was obtained as a white solid (0.90 g, 60%): [R]25
(c 1.2, CH2Cl2); IR (KBr) 2944, 2867, 1733 cm-1; H NMR (400
1
D
-22.2 (c 2.4, CH2Cl2); IR (KBr) 3304, 3081, 1687, 1593 cm-1
;
MHz, CD3OD) δ 7.43 (d, J ) 1.3 Hz, 6H), 7.25 (t, J ) 7.2 Hz,
6H), 7.17 (t, J ) 7.2 Hz, 3H), 4.68 (m, 1H), 3.82 (m, 4H), 3.32
(m, 2H), 2.01 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 170.9, 146.1,
128.5, 128.3, 126.9, 88.4, 78.6, 70.5, 65.3, 51.6, 21.0; MS (EI)
m/z 244 (+HCPh3) (100), 183, 165, 105, 77; HRMS (FAB) calcd
for C25H26NO3 (M+ + H) m/z 388.1913, found 388.1934.
(2S,3S)-Acetic Acid 3-tert-Butoxycarbonylaminoxetan-2-yl-
methyl Ester (23). (2S,3S)-Acetic acid 3-(tritylamino)oxetan-2-
ylmethyl ester (22) (120 mg, 0.31 mmol) was dissolved in dry
CH2Cl2 (10 mL) and MeOH (0.05 mL) under N2 at rt. A solution
of trifluoroacetic acid (0.1 mL) in CH2Cl2 (3 mL) was added via
syringe in 0.5 mL portions until the starting material was consumed
(30-120 min). Triethylamine (0.44 mL, 3.1 mmol) was diluted in
CH2Cl2 (3 mL) and added slowly over 10 min. Di-tert-butoxycar-
bonyl dicarbonate (0.20 g, 0.94 mmol) was added, and stirring was
continued overnight. Evaporation of all solvents yielded a yellow
oil that was purified by flash column chromatography on silica gel
(EtOAc/petroleum ether 30:70). The resulting Boc-protected acetate
23 was obtained as a clear oil (60 mg, 79%): [R]25D +15.4 (c 1.0,
1H NMR (400 MHz, CDCl3) δ 7.47 (m, 6H), 7.15-7.29 (m, 9H),
4.42 (m, 1H), 4.12 (m, 1H), 3.95, (m, 1H), 3.88 (dd, J ) 5.7, 5.7
Hz, 1H), 3.49 (dd, J ) 5.5, 5.5 Hz, 1H), 2.50 (d, J ) 11.7 Hz,
1H); 13C NMR (100 MHz, CDCl3) δ 170.9, 146.1, 128.9, 128.5,
126.9, 80.2, 79.5, 70.3, 55.3; MS (EI) m/z 243 (+CPh3) (100), 228,
215, 165, 77; HRMS (FAB) calcd for C23H22NO (M+ + H) m/z
328.1701, found 328.1714.
3-(Tritylamino)-1,5-dioxaspiro[3.2]hexanes (17). (S)-2-Meth-
ylene-3-(tritylamino)oxetane (16) (0.26 g, 0.78 mmol) was dissolved
in CH2Cl2 (20 mL), and the solution was cooled to -78 °C under
N2. A solution of dimethyldioxirane (0.35 M in CH2Cl2, 2.7 mL)
in CH2Cl2 was added in two portions over 15 min. The reaction
was stirred at -78 °C for an additional 1 h, slowly allowed to warm
to room temperature, and concentrated in vacuo. The resulting 2:1
mixture of diastereomers (17) was isolated as a white foam and
used without further purification (0.25 g, 90%): IR (KBr) 3056,
2926, 1725, 1491, 1447 cm-1: major diastereomer 1H NMR (400
MHz, CDCl3) δ 7.43 (d, J ) 7.4 Hz, 6H), 7.22 (m, 9H), 4.45-
520 J. Org. Chem., Vol. 73, No. 2, 2008

Synthesis of epi-Oxetin
1
CH2Cl2); IR (KBr) 3343, 1741, 1699, 1522 cm-1; H NMR (400
MHz, CDCl3) 5.13 (d, J ) 6.1 Hz, 1H), 4.72 (m, 3H), 4.38 (m,
2H), 4.19 (m, 1H), 2.11 (s, 3H), 1.44, (s, 9H); 13C NMR (100 MHz,
CDCl3) δ 171.0, 154.8, 86.9, 80.4, 75.3, 65.2, 47.2, 28.5, 21.0;
MS (EI) m/z 172, 159, 116, 87, 57 (100). Anal. Calcd for C11H19-
NO5: C, 53.87; H, 7.81; N, 5.71. Found: C, 53.92; H, 7.85; N,
5.64.
30%): [R]25 -61.2 (c 0.5, CDCl3); IR (KBr) 3423 (br), 2918,
D
1627, 1466, 1000 cm -1; H NMR (400 MHz, CDCl3) δ 5.34 (br
1
s, 1H), 5.10 (d, J ) 5.5 Hz, 1H), 4.80 (m, 1H), 4.70 (m, 1H), 4.58
(dd J ) 6.0, 5.7 Hz, 1H), 1.34 (s, 9H); 13C NMR (100 MHz, CDCl3)
δ 171.0, 156.5, 84.0, 82.6, 73.2, 49.3, 28.4; (EI) m/z 116, 87, 57
(100); HRMS (FAB) calcd for C9H16NO5 (M+ + H) m/z 218.1028,
found 218.1037.
(2S,3S)-3-tert-Butoxycarbonylamino-2-hydroxymethyloxet-
ane (24). (2S,3S)-Acetic acid 3-tert-butoxycarbonylaminoxetan-2-
ylmethyl ester (23) (0.15 g, 0.61 mmol) was dissolved in CH2Cl2
(5 mL). A solution of NaOMe in MeOH (0.5 M, 0.18 mmol, 0.37
mL) was added via syringe, and the solution was stirred under N2
until the starting material was consumed (∼30 min). Dowex resin
(cation exchange resin, C-211, H+ form) was added until the
solution reached a pH of 7. The resin was filtered, and the remaining
solvent was evaporated in vacuo. Purfication by flash column
chromatography on silica gel (EtOAc/petroleum ether 50:50) yielded
(2S,3S)-Aminooxetane-2-carboxylic Acid (26). (2S,3S)-tert-
Butoxycarbonylaminooxetane-2-carboxylic acid (25) (0.036 g, 0.17
mmol) was added to dry CH2Cl2 (2.5 mL) under N2. Trifluoroacetic
acid (0.2 mL) was added, and the solution turned dark yellow. After
the solution was stirred for 3 h, all solvents were removed, and the
resultant orange oil was purified on a column of Dowex ion-
exchange resin 50WX2-400. The column was washed with water
and eluted with 0.5 M aqueous ammonia to yield epi-oxetin (26)
(12 mg, 63%) as a pale yellow oil: [R]25D -13.6 (c 1.0, H2O) [lit.2
[R]25D -12.0 (c 1.0, H2O) ]; 1H NMR (300 MHz, D2O, 1,4-dioxane
as an internal standard) δ 4.98 (d, J ) 5.7 Hz, 1H), 4.76 (dd, J )
alcohol 24 (0.12 g, 81%) as a clear oil: [R]25 -5.2 (c 1.0, CH2-
D
1
Cl2); IR (KBr) 3374 (br), 2990, 1720, 1516 cm-1; H NMR (400
8.0, 8.0 Hz, 1H), 4.57 (dd, J ) 6.0, 7.5 Hz, 1H), 4.27 (m, 1H); 13
C
MHz, CDCl3) δ 5.21 (br s, NH), 4.65, (m, 3H), 4.44 (dd, J ) 6.4,
6.4 Hz), 3.83 (dd, J ) 3.7, 12.2 Hz, 1H), 3.77 (m, 1H), 2.98 (br s,
1H), 1.46 (s, 9H); 13C NMR (100 MHz, CDCl3) δ 155.3, 90.0,
80.6, 74.4, 63.7, 47.4, 28.5; MS (EI) m/z 117, 87, 73, 57 (100);
HRMS (FAB) calcd for C9H18NO4 (M+ + H) m/z 204.1236, found
204.1227.
(2S,3S)-tert-Butoxycarbonylaminooxetane-2-carboxylic Acid
(25). (2S,3S)-3-tert-Butoxycarbonylamino-2-hydroxymethyloxetane
(24) (0.067 g, 0.33 mmol) was added to a biphasic solution of CCl4
(1.0 mL), CH3CN (1.0 mL), and H2O (1.5 mL) at rt. Sodium
metaperiodate (0.29 g, 1.4 mmol) was added along with a catalytic
amount of ruthenium trichloride hydrate (2 mg, 2 mol %) upon
which the biphasic solution turned red. After 2 h, CH2Cl2 (10 mL)
was added, and the phases were separated. The aqueous phase was
then extracted with CH2Cl2 (3 × 10 mL), and the combined organic
extracts were dried (MgSO4), filtered, and concentrated. The residue
was redissolved in ethyl acetate, filtered through a pad of celite to
remove gray solids, and concentrated. The resulting crude oil was
purified via flash column chromatography on silica gel (CH2Cl2/
MeOH 98:2), providing carboxylic acid 25 as a clear oil (20 mg,
NMR (75 MHz, D2O, 1,4-dioxane as an internal standard) δ 176.5,
83.5, 72.1, 49.5.
Acknowledgment. Leon Ghosez (Universite´ Catholique de
Louvain and Institut Europe´an de Chimie et Biologie) provided
insightful suggestions on the trityl protecting group. Matt Dunn
conducted some preliminary studies on the reactivity of the Boc-
serine series. Rob Liskamp (Utrecht University) is acknowledged
for helpful discussion on the preparation of N-tritylserine. Dr.
Martha Morton is acknowledged for assistance with NMR
experiments. This paper is based upon work partially supported
by the National Science Foundation (NSF) under Grant No.
CHE-0111522.
Supporting Information Available: Copies of high-resolution
1H and 13C NMR spectra for those new compounds for which
elemental analyses are not reported. This material is available free
of charge via the Internet at http://pubs.acs.org.
JO7018762
J. Org. Chem, Vol. 73, No. 2, 2008 521